CNS*2020 Online has ended
Welcome to the Sched instance for CNS*2020 Online! Please read the instruction document on detailed information on CNS*2020.
Back To Schedule
Sunday, July 19 • 7:00pm - 8:00pm
P115: Approximating Information Filtering of a Two-stage Neural System

Log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
Google Meet link: meet.google.com/zmy-cdqa-gea
If you miss the presentation or have further questions, I would be more than happy to be contacted at: gregory.knoll[at]bccn-berlin[dot]de

Download the poster
Read the paper in Biological Cybernetics

Gregory Knoll, Žiga Bostner, Benjamin Lindner 

Information streams are processed in the brain by populations of neurons tuned to perform specific computations, the results of which are forwarded to subsequent processing stages. Building on theoretical results for the behavior of single neurons and populations, we investigate the extent to which a postsynaptic cell (PSC) can detect the information present in the output stream of a population which has encoded a signal. In this two-stage system, the population is a simple feedforward network of integrate-and-fire neurons which integrate and relay the signal, reminiscent of auditory or electroreceptor afferents in the sensory periphery. Depending on the application, the information relevant for the PSC may be contained in a specific frequency band of the stimulus, requiring the PSC to properly tune its information encoding to that band (information filtering). In the specific setup studied here, information filtering is associated with detecting synchronous activity. It was found that synchronous activity of a neural population selectively encodes information about high-frequency bands of a broadband stimulus, and it was hypothesized that this information can be read out by coincidence detector cells that are activated only by synchronous input. Firstly, we test this hypothesis and match the key characteristics of information filtering, the spectral coherence function, of the PSC and the stimulus and of the time-dependent synchrony in the population output and the stimulus; we show that the relations between the synchrony and PSC thresholds and between the synchrony window and PSC time constant are roughly linear, which implies that the synchronous output of the population can be taken as a proxy for the postsynaptic coincidence detector and, conversely, that the PSC can be made to detect synchrony (or coincidence) by adjusting its time constant and threshold. Secondly, we develop an analytical approximation for the coherence function of the PSC and the stimulus and demonstrate its accuracy by comparison against numerical simulations, both in the fluctuation-dominated and mean-driven regimes of the PSC.

avatar for Gregory Knoll

Gregory Knoll

Physics, Humboldt-Universitaet zu Berlin
B.S., Biopsychology, UC Santa BarbaraB.E., Computer Engineering, City University of New YorkM.S., Computational Neuroscience, BCCN BerlinCurrently pursuing a doctorate in the lab of Professor Benjamin Lindner at Humboldt Universität zu Berlin

Sunday July 19, 2020 7:00pm - 8:00pm CEST
Slot 08